Association of Research Libraries (ARL®)

Scholars and historians tell us that technological change is a powerful engine driving economic and social change. Invent the automobile, they argue, and suburbs are inevitable, not to mention a sexual revolution and jokes about the Yugo. This principle has penetrated the public consciousness so far that now every technical innovation is greeted with enthusiasm by those who see ways for this gadget to shape the world to suit their particular tastes.

But the distance between hope and realization is great, and there are plenty of opportunities for imagination and ineptitude to compete with each other to shape the details of a new landscape. This is a book about hope and imagination in one corner of the emerging landscape of cyberspace. It embraces passionate discussion of an idea for taking to the Internet to revolutionize one piece of the world of publishing.

"Publishing" is a business with many faces, held together in the mind's eye by the common technologies of printing and binding. Comic books, phone books, popular novels, and scientific texts all find their audiences through one facet or another of an industry whose business decisions are made on certain common grounds: costs of acquiring the content and preparing it for publication, costs of mechanical reproduction, and costs of distributing it effectively to the target audience. Those costs are recovered in many ways and even profits are sometimes made: advertisers pay most of the cost of newspapers, end users pay for best sellers, and university budgets pay for much scientific literature both by contributing "page charges" to get faculty articles published and by paying for the journals in which they are published. But despite those diversities, the common technology gives all such "publishing" a common language for evaluating continuing forms of publishing and proposed innovations.

Not every branch of "publishing" is equally healthy in the late twentieth century. The romance novel is robustly successful, but the scientific journal and the scholarly monograph are threatened by rising costs, rising output, and constrained academic budgets. The most painful paradox is that in the interests of science, the law of the market cannot be allowed to function. An item with a very small market may yet be the indispensable link in a chain of research that leads to a result of high social value. Worse, even that value may be -- at the moment of production of the small market item -- years or even decades away. The research libraries of the world have the unenviable task of seeking in common to acquire the totality of scientific and academic discourse for researchers to use, even as that totality explodes in size and cost.

Enter technology. The commonality of printing and binding is less and less necessary to transmit information to a wide audience. Some parts of print publishing have already been vulnerable to this, as newspapers have seen their markets erode under the impact of radio and television, even while odd synergies emerge and the newspaper becomes the place to find a print record of the coming week's television programs. No linear connection can be counted on: film and television entertain the eyes and occasionally the mind, but bookstores are fuller than ever of diverting books and magazines.

Accordingly, it is impossible to say with any confidence what the introduction and widespread use of networked electronic technology will do to "publishing". It is easy enough to look at the power of the World-Wide Web to transmit words far and wide and conclude that there will be some impact, but where that impact will be felt is another question. What will we read on-line? What will still need paper? To look ahead in this way is to begin to see a world in which things commonly thought of together begin to disresemble each other more than they do now. The telephone book, for example, is far less efficient than an on-line index (with yellow pages, cross-references, reverse indexes, searching-by-address, and especially a click-and-dial feature) would be, and we may confidently imagine not only that the annual massacre of trees to produce mountains of new phone books will end, but that soon we will even forget to think of the information as the phone "book" in any meaningful sense. Even as we are all reading our Jane Austen on a summer's day in a hammock twenty years from now, the "book" will have changed by virtue of the things that won't be in book form any longer. Where will the scholars and scientists be?

The pages that follow are a fragment of serious and urgent discussions that are going on at this moment in many parts of the world where people worry about the future of scholarship and science. If -- the instinctively framed argument goes -- traditional academic publishing is caught in a price spiral that threatens its stability and if a new, powerful technology that wipes out whole sectors of the costs of distributing information is now available, won't it make sense to transport as much as we can of what scholars and scientists publish to the new environment as rapidly as possible? Won't that solve the economic problems of publishing?

Perhaps, perhaps not. One thing that has emerged in the last four years or so of serious discussion on these topics is that there are various competing interests at stake. All profess to have the good of the higher educational and scientific mission in mind, but it cannot be denied that everyone seeks a future in which not only that mission but also their own economic and professional future is assured. Professors, researchers, librarians, commercial publishers, learned societies, and university presses have all emerged with distinctive points of view. The discussions are not empty of practical consequences, for strong ideas well expressed are the basis for the confidence people will place in experiments and projects that go beyond the experimental in bringing a vision of electronic publishing for the scientific and academic community to reality. Many such experiments are already well under way, and indeed we reach a moment at which people are not so much impatient to try things as fidgety that we are not yet moving past experiments to full-scale realization of ideas. Yet those fidgets may be premature, with both the results of the experiments and the convincing formulation of the theory still up in the air. The "killer application" of electronic publishing seems not to have emerged yet.

The discussions now under way are taking place in a variety of forums. They range from the informality of the departmental coffee room to full-scale conferences, some of which resemble summit meetings between mistrustful opponents. Books and articles appear apace, the traditional form of publication. But between all these is discussion facilitated by the new media of communication themselves. It is no accident that the Internet is a buzzword of this decade, for the land rush of settlers to that frontier of cyberspace has been of majestic proportions, far outstripping any recorded human settlement on planet earth. No one knows quite what to make of the numbers, but it is beyond denying that many millions of people have staked a claim to a little piece of this curious world where you can go to live without giving up the mortgage or the lease back in ordinary reality. And the one thing people do when they move to cyberspace is talk to each other.

They talk to each other there with the promiscuity and eagerness of -- well, of people. Those who read what they find on the Internet with the skeptical eye of a consumer of print publications are often dismayed, even appalled by what they see, forgetting that human beings as a rule speak to one another in ways raw, unpolished, even illogical, far below the standards of the written word. It is not so much that cyberspace is begetting trash as that it is giving a degree of publicity and visibility to our ordinary, everyday conversations, and we are -- as representatives of the species -- embarrassed by what we see.

But the caricature of Internet discussion must not blind us to the recognition that just as in everyday life, people do rise to higher levels of coherence and lucidity. The constant surprise for regular visitors to the Internet is just how many interesting people are found amid the dullards and how much stimulating and otherwise impossible conversation goes on there.

So this book is an intersection between the issues of the time and the media that make those issues arise. It is a discussion about using the networks to publish, and it is possible only because the network was there in the summer of 1994 to make this feasible.

It began with the "subversive proposal" of the title. The author of that provocative document is a pioneering veteran of network culture. Stevan Harnad, who took up a position as Professor of Psychology at the University of Southampton (UK) in the fall of 1994 after many years as a researcher and editor in Princeton, New Jersey, knows the twin worlds of paper and electronic scholarly journals as few others. He has been the editor for many years of Behavioral and Brain Sciences, a journal published by Cambridge University Press, and in 1990 he introduced Psycoloquy, the first peer-reviewed scientific journal on the Internet. He is particularly known (apart from his substantial scientific work) for his interest in the mechanisms of "peer review", by which the quality of published scientific research is guaranteed. He has pioneered in cyberspace the technique practiced by some paper journals of "open peer review", whereby the experts consulted on an article's worthiness submit their comments for publication along with the article, thus stimulating discussion in a way the article alone could not. Psycoloquy has been a notable success in achieving and maintaining a high standard of quality and in its ability to assure its audience of that quality.

Harnad has been a strong proponent of the use of the electronic networks to extend the reach and power and accessibility of scientific information, and he has not been slow to point out the economic advantages as well. Crudely put, it seems that the publisher is superfluous. If the originating scientist who does the research has access to the network, then he or she can "publish" with a few keystrokes; and anyone with network access anywhere in the world can get the results in a matter of minutes. No money need change hands.

That oversimplifies the position Harnad takes in his "subversive proposal", but it catches the gist of the unspoken hope or assumption or dream of many who look at this new environment. Harnad fully recognizes that the network itself and the tools individuals use to access it do not come free of charge, but once we can assume the ubiquity of those tools (and in the scientific community they are very nearly ubiquitous) their use adds marginal costs that are at most trivial. (The best current guess about the cost for using the Internet superhighway itself is about a dollar a year per user: Hal Varian of the University of Michigan studies this topic in ARL's Filling the Pipeline and Paying the Piper [1995].) How can we take advantage of this?

The simplest answer is that the problem is not technical, but sociological. How do we get the right user community interested and committed to communicating in this way? The first step is to talk up the idea, talk it through, and talk it out to the point of practical application.

This vital step was taken in the summer of 1994 in the discussion presented here. Harnad began it all on June 27th by posting his "subversive proposal" to the discussion list VPIEJ-L. (In the alphabet soup of the Internet, that is the name of a discussion list based at Virginia Polytechnic Institute, a particularly venturesome and interesting academic base for network activity; the list is devoted to "electronic journals", of which there are already hundreds, many dozens of them quite scholarly. ARL's Directory of Electronic Journals, Newsletters, and Academic Discussion Lists is appearing [1995] in its fifth edition.) A great deal of the conversatoin was also branched to SERIALST@uvmvm.uvm.edu, a discussion list for serials librarians and other members of the serials "chain." Some of it was also copied to other lists on the Internet, BITNET, and Usenet.

The response to Harnad's proposal (it is the opening section of this book) was quick, lively, and wide-ranging. As is often the case, the most enthusiastic discussants were those who shared much of Harnad's position. One principal discussant is Paul Ginsparg, research physicist at Los Alamos National Laboratory and a recognized innovator in electronic scientific preprint publishing as well. Ginsparg runs a service that provides, free of charge to users, the full text of "preprints" reporting the most current scientific activity in high energy physics to readers around the world. Such preprints have been in use as a paper artifact among scientists for many years; in high energy physics, the paper circulation of such things has ended, and Ginsparg's service is the faster, easier, more widely circulated result. That server exemplifies the "free" use of the Internet for high quality information distribution. To be sure, the field of high energy physics is well defined by the laboratories where work gets done, and there is a kind of quality control exercised by readers who note where given papers come from and choose how to apportion their reading time appropriately. That is a facility that would not be so easily obtained if the field were, say, nineteenth-century English literature, where a small state college may produce work as reliable as large, well-endowed institutions do.

Also contributing sympathetically to much of the discussion is Andrew Odlyzko, research scientist at AT&T Bell Labs and himself author of a forthcoming article reprinted here in abridged form that makes arguments similar to Harnad's.

But discussion was not one-sided. There are serious objections to be made and considered to what Harnad proposes. If a large and diverse body of authors produce material and a large and diverse body of readers come looking for it, it is far from obvious that the match of author to reader will be easy and transparent. The nagging question for many Internet services today is "Does it scale up?" If the whole world does this, can we still afford to do it so cheaply? What kind of editing, peer reviewing, cataloguing, and finding services will the reading public, however scholarly, find necessary? How expensive will they be? Will we migrate to a new environment, only to find that we have reinvented there all the things that cost so much in the old environment? If so, what will we have gained?

Representatives of these caution-giving points of view are strongly heard here. Two in particular deserve mention. Lorrin Garson manages the publishing operations of the American Chemical Society, a huge and vital link in the scientific information chain. He does not represent a profiteering private corporation seeking to maximize return on investment, but he has real payrolls to meet and real expenses. He speaks vigorously for the value of the middle-man's contribution to the process of scientific communication and hence for the costs that will still be incurred. Garson's viewpoint is seconded by Richard Entlich, a Librarian at Cornell University, whose libraries have cooperated with the ACS and other electronic publishing experimenters in making electronic informaiton available to users. From a European perspective, Bernard Naylor, University Librarian at the University of Southampton, weighs in from the vantage point of one who must manage the point of contact between a wider audience and the universe of scholarly and scientific discourse as a whole. The value of this whole discussion lies in the vigor and frankness with which these various participants make their cases, while maintaining a remarkably cordial and collegial tone. As is the way of the Internet, other participants came and went, making their contributions.

Here we should emphasize the point to which Harnad returns with most frequency. He speaks from the first lines of his proposal of "esoteric" publishing, that is to say, work done by the few for the few. While some participants worry over the appropriateness of the exact word Harnad prefers, all are agreed that the exchange of information among researchers working at a high level is a different matter from flooding the world with romance novels. It is central to this discussion to keep that restriction in mind and to see what difference it makes. If the "phone book" disappears from the world of "publishing" in one direction, it can be argued that the scientific journal, which few if any ever read in a hammock, will disappear in another direction, replaced by a better, faster, more reliable form of communication.

The bulk of the discussion lasted about two months. But it did not disappear. For months afterwards, conversations among real living breathing people face to face and by telephone as well as e-conversations on the more ethereal medium of the network continued to refer back to "the subversive proposal discussion". The principals in the discussion still continue to maintain the discussion among themselves, copying their messages to a few other people who have expressed the most interest in being involved. The final section of this book, on citations of articles, represents that portion of discussion which was continuing as this book went to press.

It became clear that significant issues had received here the canonical form of presentation by which they would be discussed for some time to come. A stage in the evolution of ideas about electronic publication had been set. This book recognizes that fact by presenting the discussion, still now in a timely enough fashion to affect debate, and as a record of how the argument was moving at a turning point in the network's history.

One way to present the material, of course, would have been to invite the participants in the conversation to write formal papers for an anthology. The book that resulted then would have been familiar to a fault, and easily enough understood. But first of all the participants are all busy scientists and professionals, whose own work continues to move on restlessly in search of the cutting edge. Not all would have been willing to go back and rehearse their arguments one more time. Second, much of substance would be lost by such a stereotyped presentation.

For the real value of networked discussion of this kind is in a flexibility that is impossible both in print and in real life. In real life, we may have conferences to discuss a subject, but if halfway through the conference some novel issue arises, expert testimony can be difficult to well-nigh impossible to summon in a timely way. But on the Internet, the next expert is only an e-mail message away, and more than once here that kind of "deus ex machina" arrives to stimulate and advance the discussion just in time.

In print, on the other hand, everybody is a monologist. A collaborative volume turns into a series of individual turns, where author A deals not with author B, but with an idiosyncratic version of author B, whereupon B deals not with A, but with a version of A. Examples of the consequent phenomenon of perfectly intelligent and responsible people arguing past each other are not far to seek. But the flow of e-mail makes high-minded neglect of the other's point of view impossible. The network facilitates dialogue. If that dialogue sometimes descends into "flaming", that is only a sign that real human beings are rubbing ideas together, a process as volatile in cyberspace as it is in real life. From a scientific point of view, the value lies in the give and take, which may either lead to agreement or to a nuanced and valuable expression of the irreducible conflicts. The latter is, if anything, more valuable than the former, for the way it reflects the honest truth about the world, that there are many many issues on which reasonable people do not agree with each other. Traditional print media, privileging one voice at a time, can make it seem as if the goal is to find the one authoritative voice speaking truth from on high, when a more rational sense of what we are as people would suggest that we are better served when we honestly admit our disagreements and try to make our world out of respect for the points where we cannot all march together.

Accordingly, we have elected to present the discussion as it happened, as a series of e-mail messages that flowed over the network through the summer of 1994. At first, this may be off-putting to those who do not know this culture, but within a page or two, readers will get used to the rhythm. In particular, note the fairness with which respondents quote their interlocutors by incorporating paragraphs of earlier messages in later ones, to put idea/response together for readers. If, after all, some hundreds of readers are seeing these messages separated by hours or days, they need some reminding of where the course of discussion has gone. This inevitably creates some repetition when the messages are brought together here, but the clarity is also welcome.

We have edited with a very light hand. Only tiny irrelevancies have been deleted, and occasionally we have "corrected" the sequence of messages to make the flow of thought more obvious. (Every message comes on the network with a date and time stamp, and that is the chronological order we have mainly followed, but those numbers can confuse when a message written in Europe is answered "before" it was written by a prompt respondent five or six time zones to the west.) We have left authors with their individual styles, including those who (in a practice common on the network but attracting very little comment that we have seen) follow the capitalization style of archy or e. e. cummings, but we have also corrected what were assuredly typos. We have broken the flow of the argument up into "chapters" and added a few explanatory notes to give readers of the print version all the advantages that contemporary readers of the original had. We have also provided a short glossary of netspeak.

The discussion itself had no formal conclusion: real life doesn't have many neat endings short of the pine box six feet under. The issues are not resolved. The last phase of the discussion was a lively look at the question of just how often scientific publications are read/cited, a topic of notorious difficulty. In retrospect, the discussion ended when people in other venues began speaking of it in the past tense. Stevan Harnad archived the e-mail in a long series of files stored on a networked server at Princeton University, which has been our authority for most of what is contained here (but we have supplemented it with some messages kept separately). It will have had some readers from that source, but few would have known where to look, much less how to interpret what they found there.

We have added a brief concluding essay that picks up some of the issues and ventures to outline some of the terms for the discussion that must now ensue, but we must emphasize that we do so without consultation with the participants. (Both editors appear in fleeting walk-on roles in these pages, but they had no influence over the shape of the discussion and were for the most part simply fascinated observers.)

And so this book you hold in your hands is a curious thing. It is a traditional print publication, freezing in time a series of fleeting e-mail messages that envision a future of publishing that goes well beyond print. Will print become obsolete? The fairest answer is that it will become as obsolete as speech was made obsolete by the written word, or as calligraphy was made obsolete by print. It is a mistake to speak of succession and replacement, and more important to think of a world in which available media grow more powerful and various and the interconnections more supple and useful. We would not publish this book if we did not think the networked discussion was important and interesting, but we would not publish this book if we did not think that the networked discussion has an importance and an interest that reach beyond the network's ability to disseminate it. We hope that the readers of these pages will agree.

One last point. Many readers will disagree with the subversive proposal and its advocates, perhaps fiercely. Some of those readers will themselves be publishers or other participants in the scholarly publishing community. We hope that the unpersuaded readers will be stimulated to think through the issues the most acutely, identify the points of disagreement, and clarify the issues at stake. If this proposal is invalid, what does it mean for the present and future that it is so urgently pressed, so widely discussed, and so fiercely believed? Is that not in itself a new and important fact about the environment in which we work? What would a truly creative response to this challenge look like?

Ann Shumelda Okerson
Association of Research Libraries

James J. O'Donnell
University of Pennsylvania

NOTE ON PRESENTATION

We have retained the flavor and the form of e-mail conversation as it takes place on the net today. Each message has a header (here stripped of some technicalities that appear on-line) with date/time, author, and subject. One or two small apparent departures from chronological order restore the actual sequence of messages whose time-stamps are confused by different time zones of origin.

It is a common practice for authors to embed parts of other authors' messages in their own in order to respond or comment. On-line, these will typically be marked by angle brackets at the left margin, which sometimes get doubled or tripled as a message containing embedded text is embedded in another and then perhaps yet another. We have followed this practice with two enhancements: all such embedded messages and (more often) parts of messages are printed in slightly reduced type (they have mainly appeared already in full earlier in the discussion), and where quotations-within-quotations occur we have occasionally marked the doubled and tripled angle brackets at the left margin with the initials of the writer being quoted to help the reader follow the line of argument. Finally, in the limited typography of the net, it is conventional to use asterisks for emphasis, and we have mainly retained them in these messages.

That said, if the reader is unfamiliar with the conventions of e-mail it will take some slight patience to get a sense of the rhythm of such conversation, but one reason for publishing this book in this form is to display an example of serious conversations on urgent issues in a novel forum with its own idioms and conventions.

Forward to Chapter I